| name | detecting-shadow-api-endpoints |
| description | Discover and inventory shadow API endpoints that operate outside documented OpenAPI/Swagger specs, using traffic analysis against API gateways (Kong, AWS API Gateway, Envoy), cloud configuration scanning, and source code repository mining for undocumented routes. Use when assessing API attack surface, auditing for forgotten test environments or deprecated API versions still running, or building an API registration governance policy. |
| domain | cybersecurity |
| subdomain | api-security |
| tags | ["api-security","shadow-apis","api-discovery","undocumented-apis","zombie-apis","api-inventory","attack-surface-management","api-governance"] |
| version | 1.0 |
| author | mahipal |
| license | Apache-2.0 |
| nist_csf | ["PR.PS-01","ID.RA-01","PR.DS-10","DE.CM-01"] |
| mitre_attack | ["T1190","T1133","T1526","T1213"] |
Detecting Shadow API Endpoints
Overview
Shadow APIs are API endpoints operating within an organization's environment that are not tracked, documented, or secured. They emerge from rapid development cycles, forgotten test environments, deprecated API versions left running, third-party integrations, or developer side projects deployed without governance. Shadow APIs bypass authentication and monitoring controls, creating hidden entry points for attackers. Studies show that up to 30% of API endpoints in large organizations are undocumented, making shadow API detection a critical component of API security posture management.
When to Use
- When investigating security incidents that require detecting shadow api endpoints
- When building detection rules or threat hunting queries for this domain
- When SOC analysts need structured procedures for this analysis type
- When validating security monitoring coverage for related attack techniques
Prerequisites
- API gateway or reverse proxy with traffic logging (Kong, AWS API Gateway, Envoy)
- Network traffic capture capability (packet broker, port mirroring)
- Access to source code repositories and CI/CD pipeline configurations
- Cloud provider access for configuration scanning (AWS, GCP, Azure)
- API documentation inventory (OpenAPI specs, Swagger docs)
- Python 3.8+ for custom discovery tooling
Detection Methods
1. Traffic Analysis and Comparison
Compare live API traffic against documented OpenAPI specifications to identify undocumented endpoints:
"""Shadow API Endpoint Detector
Compares observed API traffic patterns against documented
OpenAPI specifications to identify undocumented (shadow) endpoints.
"""
import json
import re
import yaml
import sys
from collections import defaultdict
from datetime import datetime
from typing import Dict, List, Set, Tuple, Optional
from dataclasses import dataclass, field
:
method:
path_pattern:
first_seen:
last_seen:
request_count:
source_ips: [] = field(default_factory=)
status_codes: [] = field(default_factory=)
has_auth_header: =
documented: =
:
PARAM_PATTERNS = [
(re.(), ),
(re.(), ),
(re.(), ),
]
():
.documented_endpoints: [[, ]] = ()
.discovered_endpoints: [[, ], DiscoveredEndpoint] = {}
():
(spec_path, ) f:
spec_path.endswith():
spec = json.load(f)
:
spec = yaml.safe_load(f)
paths = spec.get(, {})
path, methods paths.items():
normalized_path = re.sub(, , path)
method methods:
method.upper() (, , , , , , ):
.documented_endpoints.add((method.upper(), normalized_path))
()
() -> :
path = path.split()[]
pattern, replacement .PARAM_PATTERNS:
path = pattern.sub(replacement, path)
path
():
patterns = {
: re.(
),
:
}
(log_file, ) f:
line f:
log_format == :
:
entry = json.loads(line)
method = entry.get(, entry.get(, ))
path = entry.get(, entry.get(, ))
status = (entry.get(, entry.get(, )))
ip = entry.get(, entry.get(, ))
timestamp = entry.get(, entry.get(, ))
has_auth = (entry.get(, entry.get(, )))
json.JSONDecodeError:
:
= patterns[log_format].(line)
:
method = .group()
path = .group()
status = (.group())
ip = .group()
timestamp = .group()
has_auth = line
path.startswith() path.startswith():
normalized = .normalize_path(path)
key = (method.upper(), normalized)
key .discovered_endpoints:
.discovered_endpoints[key] = DiscoveredEndpoint(
method=method.upper(),
path_pattern=normalized,
first_seen=timestamp,
last_seen=timestamp,
request_count=,
documented=(key .documented_endpoints)
)
endpoint = .discovered_endpoints[key]
endpoint.request_count +=
endpoint.last_seen = timestamp
endpoint.source_ips.add(ip)
endpoint.status_codes.add(status)
has_auth:
endpoint.has_auth_header =
() -> [DiscoveredEndpoint]:
shadows = []
key, endpoint .discovered_endpoints.items():
endpoint.documented:
shadows.append(endpoint)
shadows.sort(key= e: e.request_count, reverse=)
shadows
() -> :
risk_score =
endpoint.has_auth_header:
risk_score +=
endpoint.request_count > :
risk_score +=
endpoint.request_count > :
risk_score +=
(endpoint.source_ips) > :
risk_score +=
endpoint.status_codes endpoint.status_codes:
risk_score +=
endpoint.method (, , , ):
risk_score +=
sensitive_patterns = [, , , , ,
, , , , ]
pattern sensitive_patterns:
pattern endpoint.path_pattern.lower():
risk_score +=
risk_score >= :
risk_score >= :
risk_score >= :
() -> :
shadows = .identify_shadow_apis()
total_documented = (.documented_endpoints)
total_discovered = (.discovered_endpoints)
report = {
: datetime.now().isoformat(),
: {
: total_documented,
: total_discovered,
: (shadows),
: ,
},
: []
}
endpoint shadows:
risk = .classify_risk(endpoint)
report[].append({
: endpoint.method,
: endpoint.path_pattern,
: risk,
: endpoint.request_count,
: (endpoint.source_ips),
: endpoint.has_auth_header,
: (endpoint.status_codes),
: endpoint.first_seen,
: endpoint.last_seen,
})
report
():
detector = ShadowAPIDetector()
spec_files = sys.argv[:] (sys.argv) > []
spec spec_files:
spec.endswith((, , )):
detector.load_openapi_spec(spec)
detector.process_access_log()
report = detector.generate_report()
()
()
()
()
()
()
()
ep report[]:
risk_marker = {: , : , : , : }
()
()
(, ) f:
json.dump(report, f, indent=, default=)
()
__name__ == :
main()
